Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method comprising: detecting a disaster at a primary data center, the primary data center comprising an indexing and search manager comprising an index core and a search core, a distributed file system, and a distributed storage system, the distributed file system and distributed storage system utilized in indexing of documents to be made available for search via the search core; in response to the detecting of the disaster, switching over from the primary data center to a secondary data center such that searches from one or more client applications are routed to the secondary data center, the secondary data center comprising an indexing and search manager comprising an index core and a search core, a distributed file system, and a distributed storage system, the search core and distributed file system updated periodically via duplication of the search core and distributed file system from the primary data center; for each document stored in the search core of the secondary data center: requesting a count for the document from a first client application; determining whether the count for the document from the first client application matches a count for the document from the search core of the secondary data center; and in response to a determination that the count for the document from the first client application does not match a count for the document from the search core of the secondary data center, requesting a full publish for the document from the first client application; wherein the method further comprises performing the requesting the count, the determining, and the requesting of the full publish for each document for each tenant represented in the search core of the secondary data center.
A system recovers from a primary data center disaster by switching search queries to a secondary data center. Both data centers have an indexing and search manager, including index and search cores, distributed file system, and distributed storage. The secondary data center's search core is periodically updated from the primary. After a disaster, the system checks each document in the secondary search core. It requests a "count" for the document from a client application and compares it to the count in the secondary search core. If the counts mismatch, it requests a "full publish" (re-index) of the document from the client, ensuring data consistency. This process is repeated for each document and each "tenant" (user or organization) present in the secondary search core.
2. The method of claim 1 , wherein the indexing of documents comprises: receiving primary data from a first tenant in a computer network; receiving auxiliary data relating to the primary data from the first tenant; creating a first index for the primary data and the auxiliary data from the first tenant; and storing the first index in a first shard.
The method for disaster recovery, as described previously, also includes how documents are initially indexed. Indexing involves receiving primary data (e.g., product descriptions) and auxiliary data (e.g., pricing, availability) from a "tenant" on a network. The system then creates an index combining both types of data and stores this index in a "shard," which is a partition of the overall index.
3. The method of claim 2 , wherein the first shard comprises a plurality of shardlets, each shardlet corresponding to a piece of the primary data.
The indexing method previously described organizes each "shard" into smaller units called "shardlets." Each shardlet corresponds to a specific piece or subset of the primary data that has been indexed. This enables finer-grained management and distribution of the index.
4. The method of claim 1 , wherein each document is a catalog of items and the count for a document is a count of a number of items in the corresponding catalog.
The method for disaster recovery, as described previously, defines each "document" as a catalog of items. In this context, the "count" requested and compared is the number of items within that catalog. Therefore, the system verifies the item count of a catalog in the secondary data center against the item count reported by the client application.
5. The method of claim 1 , further comprising organizing the first shard with a plurality of additional shards into the search core of the primary data center.
The method for disaster recovery, as described previously, includes a process of organizing the first shard, along with other shards, into the primary data center's search core. Therefore, the newly created index shard is integrated with existing index shards to form the complete search index.
6. The method of claim 1 , wherein the distributed storage system of the primary data center stores parsed data utilized during the indexing.
The method for disaster recovery, as described previously, stores parsed data in the primary data center's distributed storage system. This parsed data is generated during the indexing process and subsequently used to create the searchable index.
7. A system comprising: a secondary data center comprising: one or more processors; a search and indexing manager comprising: an index core; and a search core; a distributed file system; and a distributed storage system; the search and indexing manager of the secondary data center configured to: detect a disaster at a primary data center and, in response to the detecting, for each document stored in the search core of the secondary data center: request a count for the document from a first client application; determine whether the count for the document from the first client application matches a count for the document from the search core of the secondary data center; and in response to a determination that the count for the document from the first client application does not match a count for the document from the search core of the secondary data center, request a full publish for the document from the first client application, wherein the requesting the count, the determining, and the requesting of the full publish are performed for each document for each tenant represented in the search core of the secondary data center.
A system for disaster recovery in search applications includes a secondary data center with processors, a search and indexing manager (with index and search cores, distributed file system and storage). The system detects a disaster at a primary data center. In response, for each document in the secondary search core, the system: 1) requests a "count" from a client application; 2) checks if this count matches the count in the secondary search core; and 3) if the counts differ, requests a "full publish" of the document from the client, ensuring data consistency. This validation is performed for each document and each "tenant" represented in the secondary search core.
8. The system of claim 7 , wherein an index core of the primary data center and the index core of the secondary data center are configured to perform the following operations: receiving primary data from a first tenant in a computer network; receiving auxiliary data relating to the primary data from the first tenant; creating a first index for the primary data and the auxiliary data from the first tenant; and storing the first index in a first shard.
The disaster recovery system described previously includes index cores in both primary and secondary data centers. These index cores receive primary (e.g., product details) and auxiliary data (e.g., price, availability) from a "tenant". They create an index that combines both data types and stores that combined index into a "shard", which is a segment of the overall index.
9. The system of claim 8 , wherein the first shard comprises a plurality of shardlets, each shardlet corresponding to a piece of the primary data.
In the disaster recovery system described previously, each index "shard" is further divided into smaller units known as "shardlets". Each shardlet represents a smaller piece or section of the primary data that is stored inside the shard.
10. The system of claim 7 , wherein each document is a catalog of items and the count for a document is a count of a number of items in the corresponding catalog.
The disaster recovery system previously described uses "documents" where each document represents a catalog of items. When a "count" is requested, the count refers to the number of items inside the catalog that represents that document.
11. The system of claim 7 , wherein the index core of the primary data center and the index core of the secondary data center are further configured to organize the first shard with a plurality of additional shards into the search core of the primary data center.
In the disaster recovery system described previously, the index cores in both the primary and secondary data centers organize the initial shard and the subsequent shards into the search core of the primary data center.
12. A non-transitory machine-readable storage medium comprising instructions, which when implemented by one or more machines, cause the one or more machines to perform operations comprising: detecting a disaster at a primary data center, the primary data center comprising an indexing and search manager comprising an index core and a search core, a distributed file system, and a distributed storage system, the distributed file system and distributed storage system utilized in indexing of documents to be made available for search via the search core; in response to the detecting of the disaster, switching over from the primary data center to a secondary data center such that searches from one or more client applications are routed to the secondary data center, the secondary data center comprising an indexing and search manager comprising an index core and a search core, a distributed file system, and a distributed storage system, the search core and distributed file system updated periodically via duplication of the search core and distributed file system from the primary data center; for each document stored in the search core of the secondary data center: requesting a count for the document from a first client application; determining whether the count for the document from the first client application matches a count for the document from the search core of the secondary data center; and in response to a determination that the count for the document from the first client application does not match a count for the document from the search core of the secondary data center, requesting a full publish for the document from the first client application; wherein the instructions further comprise performing the requesting the count, the determining, and the requesting of the full publish for each document for each tenant represented in the search core of the secondary data center.
A computer-readable storage medium stores instructions for disaster recovery. When executed, the instructions cause the system to: detect a disaster at the primary data center which contains an indexing and search manager that includes both index and search cores, a distributed file system, and distributed storage. Upon detection, the system switches over to a secondary data center and search queries are rerouted. For each document, the system requests a count from a client application and verifies it with the count present in the secondary data center's search core. If there's a discrepancy, a full publish for the document is requested from the first client application. The requesting, determining, and requesting of the full publish are performed for each document for each tenant.
13. The non-transitory machine-readable storage medium of claim 12 , wherein the indexing of documents comprises: receiving primary data from a first tenant in a computer network; receiving auxiliary data relating to the primary data from the first tenant; creating a first index for the primary data and the auxiliary data from the first tenant; and storing the first index in a first shard.
The computer-readable storage medium for disaster recovery, as described previously, includes instructions for indexing documents. The indexing process involves receiving primary data (e.g., descriptions) and auxiliary data (e.g., prices) from a "tenant" on a network. An index is created from both data types and stored in a "shard".
14. The non-transitory machine-readable storage medium of claim 13 , wherein the first shard comprises a plurality of shardlets, each shardlet corresponding to a piece of the primary data.
The computer-readable storage medium for disaster recovery, as described previously, organizes each index "shard" into a set of smaller "shardlets," where each shardlet corresponds to a piece of the primary data from the indexing process.
15. The non-transitory machine-readable storage medium of claim 12 , wherein each document is a catalog of items and the count for a document is a count of a number of items in the corresponding catalog.
The computer-readable storage medium for disaster recovery, as described previously, specifies that each "document" is a catalog of items and that the "count" requested refers to the number of items within the catalog.
16. The non-transitory machine-readable storage medium of claim 12 , further comprising organizing the first shard with a plurality of additional shards into the search core of the primary data center.
The computer-readable storage medium for disaster recovery, as described previously, contains instructions to organize the initial shard with a plurality of additional shards into the search core of the primary data center.
17. The non-transitory machine-readable storage medium of claim 12 , wherein the distributed storage system of the primary data center stores parsed data utilized during the indexing.
The computer-readable storage medium for disaster recovery, as described previously, uses the distributed storage system of the primary data center to store parsed data which is used during the indexing process.
Unknown
November 14, 2017
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.